[187289] in North American Network Operators' Group
Re: Cisco CMTS SNMP OID's
daemon@ATHENA.MIT.EDU (Nikos Mouat)
Tue Jan 26 14:44:34 2016
X-Original-To: nanog@nanog.org
Date: Sun, 24 Jan 2016 13:47:11 -0800 (PST)
From: Nikos Mouat <nikm@cyberflunk.com>
To: Lorell Hathcock <lorell@hathcock.org>
In-Reply-To: <E8120C01-5D57-4CEF-9CA1-C103EDE6DADC@hathcock.org>
Cc: nanog@nanog.org
Errors-To: nanog-bounces@nanog.org
Hi Lorell,
Here's ones that I used to use for DOCSIS 2 UBR's years ago. For the
remote query table, you will obviously need to turn on remote query.
remote query table:
.1.3.6.1.4.1.9.10.59.1.2.1
cable modem registrations:
.1.3.6.1.2.1.10.127.1.3.3.1
if you don't have remote query capability, then you can load the cable
modem ip from the registered list, and hit it with:
transmission.127.1.2.2.1.3.2
transmission.127.1.1.4.1.5.3
which gives you transmit power and SNR from the modem, which was all that
I cared about tracking. I would keep historical logs of the values and
graph over time - but never cleared it when the modem was returned and
re-issued, which yielded interesting graphs as the modem moved from
different plants over its lifetime.
For CMTS upstream errors, I graphed the following via MRTG:
1.3.6.1.2.1.10.127.1.1.4.1.4.%d&1.3.6.1.2.1.10.127.1.1.4.1.4.%d
and for SNR:
1.3.6.1.2.1.10.127.1.1.4.1.5.%d&1.3.6.1.2.1.10.127.1.1.4.1.5.%d
with %d being the ifindex of the upstream interface(s).
Your mileage may vary - this was from my toolsets for DOCSIS 1 and DOCSIS
2 environments, and I was out of the cable business when DOCSIS 3 became
affordable for smaller operators, so never had to worry about updating
tools.
Nikos Mouat
On Sun, 24 Jan 2016, Lorell Hathcock wrote:
> All:
>
> Does anyone out there have some valuable OID's for a Cisco CMTS?
>
> The ones I am looking for are:
> Signal to Noise per upstream channel
> Cable Modem counts of all kinds
> connected / online
> ranging
> offline
>
> I opened a ticket through Cisco's help desk. I have a SmartNET contract for the unit, but they were not very helpful. The OIDs they suggested did not yield any useful data. ("0" when I know there are CMs connected, etc).
>
> Thanks in advance.
>
> Lorell Hathcock
>
>
